Method for bright and dark small targets adaptive detection
一种亮暗小目标自适应检测方法

To solve the problems of detecting bright and dark targets, an improved multi-level filter was proposed based on the method for small targets detection with traditional multi-level filter in order to preserve both bright and dark region information. Through the processing of enhancement of local contrast based on lateral inhibition for attention region, the whole bright and dark targets were enhanced with the clutter inhibited. Meanwhile the edge of the target was enhanced precisely, which was easy to compute the feature of the target. Experimental results demonstrate that the algorithm proposed can detect bright and dark small targets adaptively and effectively.